Any material, whether natural or synthetic, must exhibit satisfactory resistance to degradation under conditions of use, if products made from the materials are to find a lasting market. A lack of satisfactory resistance to degradation usually manifests itself as a partial or total loss of structural integrity, a darkening or discoloration of the product, a loss of flexibility or resilience, or a combination of the above phenomena. These phenomena are promoted or catalyzed by air (oxygen), heat and light, particularly ultraviolet light.
To protect materials, ingredients which can be collectively called stabilizers are admixed with the materials to prevent or inhibit degradation. These stabilizers work in diverse and complex ways, such that a compound which stabilizes against heat and oxygen degradation in a material may not stabilize against light degradation in the same material, or vice versa. Furthermore, a compound which acts as a stabilizer against oxygen degradation in one type of material may be relatively inactive in another type of material. Thus compounds which are stabilizers are further classed as antioxidants, antiozonants, heat stabilizers and ultraviolet (UV) light stabilizers, depending upon what type of activity and stabilization they demonstrate. In many cases, to obtain optimum protection, a mixture of compounds, each specifically selected to afford maximum protection against a certain type of degradation, is often used. In some instances stabilizers are deliberately chosen to counter the adverse effects of a plasticizer which, though highly effective as a plasticizer, tends to accelerate UV degradation. In other words, the plasticized material is more susceptible to degradation than if no plasticizer was added. As a general empirical rule, it is found that plasticizers are marginally effective as stabilizers, and stabilizers are marginally effective as plasticizers, it being more likely that a compound with desirable stabilizer properties has undesirable plasticizer properties, and vice versa.
The present invention is directed to (a) novel UV light stabilizers classed as hindered amines, more specifically classed as hindered cyclic ketodiazaalkanes, and (b) novel compositions in which the cyclic ketodiazacycloalkanes are incorporated. The basic structure of these novel compounds is a polysubstituted 2-keto-1,4-diazacycloalkane having (a) fixed two-carbon bridge between the two N atoms (the N.sup.1 and N.sup.4 atoms) of the diaza ring, the remaining portion of the ring having a variable length bridge of two or more carbon atoms, (b) an N.sup.1 -adjacent carbonyl in the fixed two-carbon bridge, and (c) at least the N.sup.4 -adjacent carbon atom of the fixed two-carbon bridge has two substituents (hence "polysubstituted"), which may be cyclizable, that is, form a cyclic substituent. These compounds which may be monocyclic, or with cyclizable substituents, may be bicyclic or tricyclic, are particularly useful as UV light stabilizers in substantially colorless organic materials. They may also form dimers and bis-compounds. The diaza ring of the basic structure may have from 6 to 9 ring members, more preferably from 6 to 8 ring members, and most preferably from 6 to 7 ring members.
It is known that 4,4,6,6-tetramethyl-1,5-diazacycloheptan-2-one may be prepared by a Schmidt's rearrangement of a six-membered ring with sodium azide (see German Pat. No. 2,428,877) but there is no known manner of similarly arriving at a six-membered 1,4-diaza ring with an N.sup.1 -adjacent carbonyl. It is also known that 1,4-diaza [3,3,5,5]dipentamethylene-2-one may be prepared, starting with cyclohexanone, by cyclization of bis(1-cyanocyclohexyl)amine, reducing with lithium aluminum hydride to form 1,4-diaza3,3,5,5-dipentamethylene-2-imino, treating with acetic anhydride and heating with hydrochloric acid. This is set out in greater detail in an article by Helmut Egg in Monatshefto fur Chemie 106, 1167-1173 (1975). However, starting with acetone instead of cyclohexanone, the reactions do not proceed in an analogous manner to give 3,3,5,5-tetramethyl-piperazin-2-one, as indicated by R. Sudo and S. Ichihara in "Bull. Che. Soc. Japan", 36, (1963). The Egg reference teaches substituted piperazines wherein each symmetrical N.sup.4 adjacent carbon is part of a six-membered ring and the cyclic substituent on each N.sup.4 adjacent carbon is always the same. A single cyclic substituent on the N.sup.4 adjacent C atom of the fixed two-carbon bridge cannot be prepared by following the techniques of Egg.
Cis-3,3-dimethyl-decahydroquinoxalin-2-one has been prepared by Bindler, J. in U.S. Pat. No. 2,920,077 from difficulty obtained cis-1,2-diaminocyclohexane and it is disclosed that the cis-compounds are valuable intermediates for the production of pharmaceuticals, textile auxiliary products and synthetic materials. This reference states that the trans-1,2-diaminocyclohexane is converted with excess chloracetic acid, or with salts thereof, into 1,2-diaminocyclohexane-n,N'-tetraacetic acid, which is quite unlike the behavior of the cis starting material. The cis-2-keto-1,4-diazacycloalkane is prepared by reacting an aqueous solution of cis-1,2-diaminocyclohexane with acetone cyanohydrin, and heating the reaction solution to dryness. The reference does not teach formation of a trans-5,6-polyalkylene-2-keto-diazocycloalkane, and there is no suggestion as to how it could be made. Nevertheless we have found that trans-2-keto-1,4-diazacyclohexane can be formed in a manner analogous to that in which the cis-2-keto-1,4-diazacyclohexane is formed.
Following the teachings of Bindler, ethylene diamine may be substituted for cyclohexanedimine and 3,3-dimethyl-2-ketopiperazine is obtained. However, when substituted ethylene diamine is used, the substituents appear on the No. 6 carbon of the diaza ring. For example, with 1,2-propane diamine, 3,3,6-trimethyl-2-ketopiperazine is formed, and with 2-methyl-1,2-propane diamine the compound obtained is 3,3,6,6-tetramethyl-2-ketopiperazine. 6-substituted and 3-substituted carbons are not symmetrical carbon atoms about the same N-adjacent atom in the diaza ring (hereinafter referred to as "symmetrical N-adjacent C atoms"). These compounds are quite unlike the novel compounds claimed. Moreover, 3,3,6,6-tetraalkyl substituted diazacycloalkan-2-ones are relatively ineffective UV stabilizers, confirming our experience that the more substituents on symmetrical N-adjacent C atoms, the better the stabilization effect.
It is known that 2,2,4-trimethyl-tetrahydroquinoline can be hydrogenated to form a mixture of cis and trans 2,2,4-trimethyldecahydroquinoline, and, in general, the trans isomer is the major constituent. However, 2,2-dimethyl-tetrahydroquinoxaline is not hydrogenated in an analogous manner. Quite unexpectedly, providing a 2-keto substituent and forming an amide which is not generally easily hydrogenated, allows the 3,3-dimethyl-tetrahydroquinoxalin-2-one to be hydrogenated to pure cis-3,3-dimethyl-decahydroquinoxalin-2-one which, when blended into a substantially colorless organic material or carrier provides a UV light absorbing composition which is stable.
It is also known that tetraalkyl substituted piperazinediones disclosed in U.S. Pat. No. 3,920,659, are useful UV light stabilizers; in these stabilizers each of two N.sup.4 -adjacent symmetrical C atoms have dialkyl substituents. These compounds can be reduced to the tetraalkyl substituted piperazine as disclosed in German Offenlegungsschrift No. 2,315,042, laid open Oct. 18, 1973. There is no suggestion, however, as to how a mono-keto structure, that is a 2-keto-1,4-diazacycloalkane structure may be prepared with a total of two or more substituents on symmetrical N.sup.4 -adjacent carbon atoms.
In addition to the tetraalkyl substituted piperazines mentioned hereinabove, it is known that tricyclo-1,4-diazaalkanes and diketo-tricyclo-1,4-diazaalkanes are good stabilizers. However, the efficacy of polysubstituted bicyclo-1,4-diazaalkanes and 2-keto-bicyclo-1,4-diazaalkanes as UV stabilizers in light absorbing, substantially colorless organic substrates, was not known, since these 1,4-diazacycloalkanes and 1,4-diazacycloalkanones were not readily available, and therefore not tested.